What do impressions, clicks and position actually mean?

Picture your page as a stall in a very large market. An impression is a shopper walking past and seeing your sign. A click is one of them stopping at your table. Everything else on the report is a way of describing those two numbers.

• Impression. Google showed your page in a set of results someone was looking at. They did not have to scroll down to it, though results sitting on a later page only count once somebody goes to that page. An impression is not a visit. Nobody has arrived yet.
• Click. Someone chose your result and came to your site. This is the only one of the four that puts a person on your page.
• Average position. Where your link sat, averaged across every appearance in the period. It is not one place your page lives. A page that sits near the top for one phrase and far down for another reports a number in the middle that matches neither.
• Click-through rate (CTR). Clicks divided by impressions, written as a percentage. Out of everyone who was shown your result, the share who chose it.

One thing worth fixing in your head early: all four describe Google Search results and nothing else. They say nothing about the assistants your customers may also be asking, which is a gap we come back to below.

• Impressions are demand you have already earned. Somebody typed something, and Google thought your page was a plausible answer. You did not buy that and you cannot fake it, which makes it the most honest signal you have about which of your pages people actually reach for.
• The gap between the two numbers is a diagnosis. A page with no impressions is not being shown. A page with impressions and no clicks is being shown and passed over. Those are different problems with different fixes, and the pair of numbers is what tells them apart.
• Clicks turn on the part of the result you actually own. What a person reads before deciding is your title and your description. The page underneath is not in that decision at all.
• Answer engines publish nothing like this. No assistant hands you an impressions report, so there is no equivalent number for the audience reading you through an AI assistant. That is exactly why this report stays useful: it is the best available read on which of your pages people already want an answer about, and those pages are the ones worth making machine-readable first.
• The phrases underneath are usually specific ones. Most of what shows up here is long-tail: full questions rather than one-word terms. That page covers the queries. This one covers the numbers they produce.

The honest part first. Google gives you all four numbers for free in Search Console, and MetaPilot is not a rank tracker and does not replace that report. If you want to browse your search performance, browse it in Google.

What MetaPilot does is use the numbers to decide what to work on and how to write it. Connect Google once, read-only, and your per-page numbers are joined to the pages MetaPilot already knows about. Here is what that buys you, and none of it is a dashboard:

• Your pages get ranked by the demand they already have. The pages earning impressions that carry no structured data are listed highest first, so the top of the list is the page that costs you the most by staying unreadable to machines.
• The queries go into the writing. When MetaPilot writes a page's title and description, the searches that page already appears for are in the prompt, instead of a keyword inferred from reading the copy.
• Read-only, always. MetaPilot never writes to your Search Console, submits sitemaps or requests indexing, and disconnecting withdraws the access at Google.

It is free on every plan, including Free. Ready to do this? Connect Search Console to MetaPilot →

It is the phrase you will see most often around this data, and it is two separate facts about one URL.

• It has impressions. Google is putting the page in front of people right now, whether or not any of them are clicking.
• It has no structured data. The HTML it serves carries no schema.org JSON-LD, so nothing on the page states in machine-readable form what it is, who publishes it, what it costs or what question it answers. A person reading it understands all of that. A machine is guessing.

Either fact on its own is ordinary. A page with structured data and no impressions is one nobody has asked for yet. A page with impressions and nothing machine-readable is the interesting one: people already arrive there, and every engine that reads it has to infer what it is about. It is usually the guide, the collection or the landing page, because the platform that generated your site covered products and stopped.

That pair is what MetaPilot's Search Console integration sorts your site by, and the reason it sorts by impressions rather than by anything else. When you are ready: Connect Search Console to MetaPilot →